With nearly 15 million units of red blood cells (RBCs) transfused to about 5 million patients in the U.S. every year, RBC transfusion is one of the most commonly prescribed therapies for hospital inpatients. In recent years, ample clinical evidence has accumulated that a significant proportion of morbidity and mortality in critically ill patientsis due to the toxic effects of RBC transfusions. Most transfusions involve RBCs that had been stored in an anticoagulant-preservative solution at 1-6 C for up to 6 weeks. The biochemical, mechanical and functional properties of RBCs deteriorate progressively - by the end of allowable storage up to 1% of stored RBCs undergo lysis, and as many as 25% of the remaining RBCs are irreparably damaged, non-viable cells. Moreover, the RBC storage medium accumulates known mediators of toxicity as byproducts of RBC metabolism and degradation. Infusion of these toxic mediators and the irreparably damaged cells into the recipient during transfusion reduces the therapeutic efficacy of transfusion and contributes to multiple adverse outcomes in 1-2% of U.S. citizens. The goal of this project is to devise a high-throughput technology for in-line removal of irreparably damaged cells and toxic mediators in the storage medium from units of stored RBCs in real-time, during the transfusion process. This project will initially focus on the design and performance optimization of the proposed technology on milliliter-size samples of stored RBCs. To support the design iterations, we will develop an auxiliary hematology-on-a-chip device for measuring all relevant geometric, mechanical and biochemical properties for thousands of individual RBCs at high throughput. At the second stage of the project, we will scale up the optimized design for processing clinical-size RBC units. We will perform a broad panel of in vitro tests to characterize the quality of processed RBCs. To perform an integrative test of the primary RBC function, we will develop another auxiliary device for measuring the ability of stored RBCs to load / offload oxygen in artificial capillary networks directly. The anticipated outcome of this project is a full-scale prototype of the proposed device that can be used to further test the processed well- preserved stored RBCs for pro-inflammatory and pro-thrombotic activity, and post-transfusion viability and intravascular survival in human subjects in vivo. Conventional paradigm postulates that all stored RBCs in a bag are homogeneous with respect to the storage-induced deterioration of their properties, and consequently attempts to 'rejuvenate' stored RBCs through manipulation of the overall storage conditions and re-formulation of additive solutions. Our approach challenges this conventional paradigm by leveraging the heterogeneity of stored RBCs to enable transfusion of only well-preserved cells, free from irreparably damaged cells and toxins in the storage medium. This is an entirely novel approach with a potentially game-changing, transformative impact on the safety and efficacy of transfusions administered throughout the practice of medicine.